Climate change and rainfall variability are driving many farmers to adopt irrigation, who were historically within rain-dependent regions, to sustain crop productivity. In the Mid-Atlantic region, irrigation in agriculture has grown since the 1980s due to rising temperatures and changes in precipitation patterns. Dry summers and uneven seasonal rainfall have necessitated a shift toward irrigation, particularly in Maryland's Coastal Plain. However, high dependence on confined groundwater for irrigation around this area has strained aquifers. To mitigate this strain, exploring alternative water sources is now important. This study investigates the viability of utilizing treated wastewater from plants as an irrigation substitute in Maryland's Coastal Plain. Using the Soil and Water Assessment Tool, the research evaluates crop productivity and irrigation needs under various climate scenarios. Results indicate that recycled water from nearby treatment plants can meet crop water requirements during wet years and partially during moderate and dry years, reducing aquifer reliance by 56 and 30%, respectively. This framework aims to boost yields while conserving freshwater resources. By serving as a decision support tool, stakeholders can assess the feasibility of recycled water for irrigation, thereby potentially reducing strain on confined aquifers.

  • This study presents a method for utilizing reclaimed water for irrigation.

  • We conducted crop yield simulations for corn and soybeans, the most dominant crops in the region.

  • Using the SWAT model, we have calculated the water requirements for corn and soybean under various climatic conditions.

  • We have determined the proportion of the total water requirement that can be sourced by water reuse.

SWAT

Soil and water assessment tool

SWAT-CUP

SWAT calibration and uncertainty programs

WWTP

Wastewater treatment plant

MCM

Million cubic meters

DEM

Digital elevation map

USGS

United States Geological Survey

SSURGO

Soil survey geographic database

CDL

Crop data layer

NLDAS-2

North American Land Data Assimilation System Phase-2

USDA

United States Department of Agriculture

NASS

National Agricultural Statistical Services

HRU

Hydrological response unit

NSE

Nash–Sutcliffe coefficients

KGE

Kling–Gupta efficiency

PBIAS

Percent bias

According to the United Nations' ‘International Decade for Action: Water for Life,’ water has emerged as a pivotal concern in the 21st century (UN Water 2015). Water usage for agricultural, industrial, and domestic purposes has multiplied more than six times in the past century. This rise, due to the needs of an expanding global population, has led to a surge in water scarcity and stress, affecting many regions worldwide (Salman 2005; Pfister et al. 2011; Kummu et al. 2016; Ritchie & Roser 2017). Agriculture is especially vulnerable to water stresses as it is highly water-dependent and climate-driven (Huang et al. 2021; Costantini et al. 2023; Reimer et al. 2023). Changes in precipitation patterns and increases in temperature, along with the reduction in the availability of groundwater sources for irrigation, negatively impact agricultural productivity and threaten global food security (Taylor et al. 2013; Lamichhane 2022; Yildiz et al. 2022; Wang et al. 2023). Excessive use of surface and groundwater resources for irrigation to maintain agricultural productivity can also have a cascading effect leading to water conflicts as well as detrimental impacts on the ecological sustainability of the region (Chiarelli et al. 2022; Fortes et al. 2022; Yang et al. 2023).

Despite its relatively small geographical footprint in the Mid-Atlantic region, Maryland is distinguished by its robust agricultural sector, which ranks as the state's primary industry. Approximately 900,000 ha within the state are allocated for agricultural production (Dong et al. 2019; Paul et al. 2021). The varied agricultural production systems across the region's diverse geography are currently facing the challenges posed by climate change (Dong et al. 2019; Teodoro & Nairn 2020; Paul et al. 2021). Climate-induced variabilities in rainfall and rising temperatures, specifically during the growing season, pose significant risks to the traditionally rainfed agriculture in the region (Paul et al. 2021). These rainfall fluctuations along with changes to seasonal rainfall patterns have led to an increase in short-term agricultural droughts (Pyke & Najjar n.d.; Teodoro & Nairn 2020) resulting in increased reliance on supplemental irrigation to maintain agricultural productivity (Shoushtarian & Negahban-Azar 2020; Paul et al. 2021). In Maryland, groundwater extraction for irrigation surged by approximately 38% between 2000 and 2015 (Paul et al. 2020). This extraction, primarily from confined aquifers, has resulted in a considerable decline in groundwater levels across the state. For instance, the Maryland Coastal Plain has experienced a groundwater level reduction of up to 23 m from 1982 to 2018 (Shirmohammadi 2019). Concurrently, the eastern United States (U.S.), particularly the Mid-Atlantic, has seen escalated irrigation withdrawals due to more frequent seasonal droughts and expansion of irrigated agriculture in the region (Reilly et al. 2008). This has led to a 15–85% rise in irrigation water demand across various Mid-Atlantic counties (Paul et al. 2021). With anticipated exacerbations from climate change, the region may confront increased water stress by 2045 (Paul et al. 2020).

Anticipated increases in irrigation demand, driven by the necessity to buffer the impacts of climate change and to satisfy increasing agricultural needs, are exerting substantial pressure on global aquifer systems, including those in the U.S. Mid-Atlantic region. Consequently, the identification of alternative irrigation water sources is imperative to enhance the sustainability of these aquifers. Furthermore, this approach is instrumental in conserving the limited freshwater resources (Urkiaga et al. 2008). Treated wastewater from municipal and industrial wastewater treatment plants (WWTPs) is one such alternative source of water for irrigation. Recycled water from the WWTPs has been utilized in the Western U.S. to supplement the traditional sources of water for supplemental irrigation to respond to the tremendous stress resulting from severe drought conditions experienced by the region (Sheikh et al. 2018; Paul et al. 2020). The use of recycled water from WWTPs for irrigation provides various benefits, including a reduction in freshwater withdrawals, easier management, recycling of nutrients, and increased water supply reliability (Shoushtarian & Negahban-Azar 2020). In addition, using the treated wastewater from WWTPs helps the well-being of the ecosystem since it is no longer discharged into the riverine systems.

The use of water from WWTPs for irrigation, however, also comes with some challenges. Quality of recycled water from the WWTPs, potential health hazards, and social acceptance factors are some of the major concerns with the use of recycled water for irrigation (Shoushtarian & Negahban-Azar 2020). It has, nonetheless, been determined that the use of recycled water from WWTPs for irrigation is not hazardous for most non-food, fiber crops, and also for food crops for which the edible portion does not come in contact with the recycled water used for irrigation (Sheikh et al. 2018; Olivieri et al. 2020). Hence, it is not necessary to use highly treated recycled water for the majority of the irrigation water use (Sheikh et al. 2018). As an example, undisinfected secondary treated wastewater is allowed for irrigation of non-food-bearing trees and fiber crops (Sheikh et al. 2018; Olivieri et al. 2020). Agricultural use of recycled water has also been successfully implemented in the central coastal region of California over the past 20 years (Sheikh et al. 2018; Olivieri et al. 2020). These successful projects can be a blueprint for successful examples of agricultural reuse projects in other regions (Sheikh et al. 2018). Currently, the rate of water reuse for agriculture in Maryland is very limited. Therefore, there is a huge potential to utilize recycled water for supplemental irrigation due to the availability of 46 WWTPs with a permitted capacity of at least 0.004 million cubic meters (MCM) per day and an additional 112 smaller WWTPs (capacity of ≤0.004 MCM per day) (Plumb 2004). As demands for water withdrawals for irrigation and municipal use are increasing due to the combined impact of climate change and growing population, it is both timely and vital to explore the full potential of using Maryland's reclaimed wastewater for irrigation to improve agricultural sustainability while conserving freshwater from confined aquifer sources in the region.

Assessing the potential for using WWTPs as an alternate source of water for irrigation requires the estimation of supplemental water required for irrigating the major crops of any region under a range of climatic conditions. The soil and water assessment tool (SWAT) is a watershed-scale, process-based model that has been widely used to simulate crop yield and estimate irrigation water use for multiple crop types under a range of agricultural management practices and climatic conditions. For example, Uniyal et al. (2019) simulated regional irrigation requirements for different agro-climatic regions across the globe using the SWAT model. The study used bias-corrected climate data as input and simulated crop yield and irrigation water requirements for the different study locations. In another study, Tibebe et al. (2016) simulated irrigation water requirements in the Holetta watershed of Ethiopia using SWAT and then compared the water demand to the total supply by surficial rivers within that watershed to compare the supply and demand of water. There are additional studies that evaluated the potential of using water from WWTPs as an alternative irrigation source. For example, Kama et al. (2023) explored the impacts of treated wastewater irrigation on agricultural systems and the food chain and Minhas et al. (2022) recognized wastewater as a valuable resource for irrigation in India. Some studies have discussed the importance of wastewater from treatment plants as a source for irrigation in the western U.S. including California (Paul 2021) and Utah (Ahmadi & Merkley 2017). The observed declines in groundwater levels and anticipated increase in the demand for supplemental irrigation in the eastern U.S. under future climate necessitates the identification and evaluation of possible alternative sources of water for irrigation in the region. To the author's knowledge, this study is the first to evaluate the prospect of WWTPs as a source of water for irrigation in southern Maryland.

This study evaluated the use of WWTPs as an alternate source of water for irrigation by performing a case study in a watershed located in southern Maryland. Currently, the region depends predominantly on rainfed irrigation strategies but is seeing an increasing demand for supplemental irrigation due to increasing climate variability in the region (Houser et al. 2023). To enhance crop yields and illustrate the potential reductions in groundwater withdrawals from confined aquifer systems, this study investigates the application of recycled water from WWTPs. The demonstrated approach can be a good example as well as an ad hoc decision support tool to evaluate the feasibility of using WWTPs for supplemental irrigation and estimate fresh groundwater conservation while also evaluating the improvement in agricultural productivity in similar regions of the Mid-Atlantic U.S. The specific objectives of this research were to: (1) determine the water requirement for irrigating the two most dominant crops (corn and soybeans) in southern Maryland under a range of climatic conditions, (2) estimate and evaluate the differences in crop yield between rainfed and irrigated scenarios for understanding the necessity of irrigating these crops for improved productivity, and (3) evaluate the feasibility of using recycled water available from nearby WWTPs to supplement the total irrigation demand and contribute to the conservation of groundwater sources of water currently being used for irrigation.

Study area

The study was conducted in the Zekiah watershed, a small headwater watershed in southern Maryland that is part of the larger Chesapeake Bay watershed and drains into the Chesapeake Bay (Figure 1).
Figure 1

Location of Zekiah Watershed in Maryland along with its land classification. SWAT-delineated subbasins are also shown.

Figure 1

Location of Zekiah Watershed in Maryland along with its land classification. SWAT-delineated subbasins are also shown.

Close modal

Forests, developed land, and agriculture are the major land uses in the watershed covering 58, 16, and 13% of the area, respectively (MDP 2010). Corn and soybeans are the two most dominant crops produced in the region, covering 3 and 3.5% area of the total land, respectively, with additional production of winter wheat, sorghum, and barley. The soil in the Zekiah watershed is characterized as moderately drained, loamy soils, with a clayey sub-soil (NRCS 1974). The watershed receives an average annual precipitation of 1,140 mm of which about 600 mm is received during the growing season from May to September (based on climate data from 1998 to 2020). The average temperature in the region ranges from −3.5 °C in the winter to 31 °C in the summer. The watershed is approximately 29 km long from the headwaters to the confluence with the Wicomico River and covers an area of about 281 km2. Freshwater use in the watershed is competed between municipal, industrial, and agricultural use (MDP 2010).

Hydrologic modeling

The SWAT model is a continuous time step and process-based hydrologic and water quality model (Arnold et al. 1998; Neitsch et al. 2011) that has been widely used for simulating the dynamic hydrological response of watersheds and the effects of best management practices (BMPs) (Williams et al. 1984; Arnold et al. 1998; Paul 2021; Paul et al. 2021). The model has also been used to evaluate the impacts of land management practices on crop yield for different crop types as well as irrigation water use for a range of climatic and geological conditions (Srinivasan et al. 2010; Narsimlu et al. 2015; Paul et al. 2021; Zare et al. 2023). Basic inputs to the model include elevation data, soil classification, land use, and weather data. Additional details including land management practices can be provided for further refinement of model inputs or for evaluating different scenarios (Srinivasan et al. 2010; Chen et al. 2018).

In the SWAT model, the watershed is divided into several subbasins, and each subbasin is further divided into hydrological response units (HRUs) based on characteristics that include land use, soil type, and slope (Arnold et al. 1998). HRUs represent the basic unit for calculation in the SWAT model. The spatial disaggregation of a watershed to subbasins and further into HRUs allows SWAT to represent the diversity of soils, topography, and land cover, thus simulating the spatial variability in hydrologic and water quality processes as well as crop yields, facilitating highly detailed spatial results.

The model uses a water balance equation to estimate different hydrological components at different spatial scales of the model (both subbasin and HRU). The hydrological components include both blue and green water. The green water includes the portion of precipitation that infiltrates and gets stored in the soil as soil moisture and then returns to the atmosphere through plant transpiration and/or direct evaporation. Bluewater, on the other hand, includes a portion of water that flows through or below the land surface and is stored in aquifers, lakes, and reservoirs (Neitsch et al. 2011; Paul et al. 2021). SWAT model uses the following water balance equation to update the daily soil moisture in the soil profile.
where SWt and SW0 are the soil water storage at time t and 0, respectively; Pday is the precipitation; Qsurf is the surface runoff flow; Ea is the actual evapotranspiration; wseep is the deep aquifer recharge; and Qgw is the groundwater flow.

SWAT simulates irrigation demand based on the soil moisture balance and ET demand. SWAT provides multiple methods for estimating ET from which Penman–Monteith was used for this study. Irrigation in SWAT can be simulated based on specific dates or by using an auto-irrigation subroutine. The auto-irrigation subroutine in SWAT triggers irrigation when the soil moisture on a given day reaches the plant water stress threshold or soil water deficit threshold set by the user (Akhavan et al. 2010; Neitsch et al. 2011). Irrigation based on plant water stress is automatically applied by SWAT when the plant stress falls below the water stress threshold value (Neitsch et al. 2011). On the other hand, irrigation based on soil water deficit is triggered when soil water in the profile falls below the field capacity by more than the soil water deficit threshold (Neitsch et al. 2011).

SWAT estimates crop yield as a product of harvest index and above-ground biomass. For crop growth, temperature is one of the most important factors with each plant type having its minimum, maximum, and optimal temperature range for growth that can be specified by the user. The heat unit theory by Boswell (1926) states that plants have heat requirements that can be quantified and linked to time to maturity and is the basis for plant growth in the SWAT model. The heat index used for simulating plant growth in SWAT is the number of degrees above the base temperature (Neitsch et al. 2011). SWAT assumes that all heat above the base temperature accelerates crop growth and maturity. The total number of heat units (degree-days) required for a plant to reach maturity is calculated by accumulating the total heat unit from the planting date to the number of days required to reach maturity (Neitsch et al. 2011). Besides irrigation and heat index, crop growth in SWAT also varies based on nutrient stresses and other management practices (e.g., fertilization, tillage, etc.). SWAT allows for scheduling management practices including planting, fertilization, irrigation, and harvest, all of which can be based on the accumulation of heat units or calendar days and are used in the model for incorporating management practices and simulating crop growth.

Model setup

Input data and model development

A 30 m × 30 m digital elevation map (DEM) data acquired from the United States Geological Survey (USGS) (https://apps.nationalmap.gov/viewer/) were used as the elevation data for model setup while soil information was acquired from the Soil Survey Geographic Database (SSURGO) (https://websoilsurvey.sc.egov.usda.gov/app) (Stermitz et al. 1999). Land use information available from the 2020 Cropland Data Layer (CDL) (https://nassgeodata.gmu.edu/CropScape/) was used as the base land use for the model (Boryan et al. 2011; Han et al. 2012).

Climate forcing data were obtained from North American Land Data Assimilation System Phase-2 (NLDAS-2) (https://ldas.gsfc.nasa.gov/nldas/), which is a comprehensive dataset that integrates observations and model simulations to provide high-quality meteorological and hydrological information over North America (Xia et al. 2012). NLDAS-2 climate variables used as model input for this study included precipitation, temperature, humidity, wind speed, solar radiation, and soil moisture and are available at an hourly time scale and a spatial resolution of 1/8th degree (Xia et al. 2012).

The SWAT model was delineated for the study watershed using a user-defined outlet near the USGS station 01660920 (Latitude 38°29′26.1″, longitude 76°55′37.5″) (Figure 1). A user-defined threshold of 5–5–15% was used for land use, soil, and slope classes, respectively, and yielded a model that included 31 subbasins and 1,224 HRUs (Figure 1). The model was set up for a simulation period of 27 years (1993–2020) that included a 5-year warm-up period.

Model calibration and validation

The SWAT model was first calibrated for crop yield (corn and soybeans) followed by calibration of streamflow. The model calibration and validation were performed using SWAT Calibration and Uncertainty Programs (SWAT-CUP), an automated calibration and uncertainty program designed specifically for the SWAT model (Abbaspour et al. 2015). The sequential uncertainties fitting version 2 (SUFI-2) (Abbaspour et al. 2004, 2007) algorithm within SWAT-CUP was used as the optimization algorithm for calibrating the model as it has been successfully used in multiple studies to calibrate and validate the SWAT model (Narsimlu et al. 2015; Kumar et al. 2017; Paul et al. 2021). In SUFI-2, the uncertainty degree is measured as the p-factor, which is the percentage of measured data bracketed by the 95% prediction uncertainty (95 PPU). Another measure is the r-factor which quantifies the strength of a calibration/uncertainty analysis by an average thickness of the 95 PPU band divided by the standard deviation of the measured data (Abbaspour et al. 1997; Khalid et al. 2016; Paul & Negahban-Azar 2018). A value of r-factor and p-value closer to 1 indicates a good model fit.

The model was evaluated for the simulation of crop yield and streamflow using graphical as well as statistical measures. Time series plots between simulated and observed variables were used to perform the graphical evaluation while the statistical evaluation was performed using the following evaluation metrics: coefficient of determination (R2), Nash–Sutcliffe coefficients (NSE), Kling–Gupta efficiency (KGE), and percent bias (PBIAS). Among them, R2 describes the degree of collinearity between the observed and simulated values (Paul & Negahban-Azar 2018). NSE is the most widely used statistic in hydrology with its value ranging from negative infinity to 1. An NSE value of 1 shows the model perfectly corresponds with the observed data while 0 shows that the predicted model is as good as the observed mean, and NSE values of less than 0 indicate that the observed mean is a better predictor than the model (Paul & Negahban-Azar 2018; Paul et al. 2021). PBIAS measures the tendency of the simulated values to be larger or smaller than their observed counterparts. Positive PBIAS shows overestimation in the model and negative value shows underestimation (Paul & Negahban-Azar 2018). KGE is based on the equal weighting of three sub-components: linear correlation, bias ratio, and variability between simulated and observed data, and a KGE value closer to 1 indicates higher model accuracy. KGE is developed because NSE is overestimated in extreme weather conditions (Paul & Negahban-Azar 2018). These evaluation metrics are discussed in detail in Paul & Negahban-Azar (2018) and Moriasi et al. (2015).

Observed county-level annual crop yield data used for model calibration and validation were collected from USDA National Agricultural Statistical Services (USDA-NASS) (https://www.nass.usda.gov/Quick_Stats/Lite) from 1998 to 2016. Crop yields for both corn and soybeans were calibrated using data from 1998 to 2007 (10 years) and validated from 2008 to 2016 (9 years). Daily streamflow data used for flow evaluation were acquired from USGS (https://waterdata.usgs.gov/nwis) from 1998 to 2013. The model was calibrated for streamflow from 1998 to 2004 (7 years) and validated from 2007 to 2013 (7 years). The years 2005–2006 had missing streamflow data and hence, were skipped from the analysis.

Scenario development

The model, after calibration and validation, was run for two irrigation management scenarios. The scenarios were developed to analyze the impact of irrigation on crop yield for corn and soybeans in the Zekiah watershed representing the Mid-Atlantic region of the U.S. where drought has been shown to have a negative impact on crop yield (MDA 2014). The scenarios were also critical to determine the amount of irrigation required to obtain optimal crop yield under different climatic conditions. This was crucial to evaluate the capability of the WWTPs to supplement the amount of irrigation required for optimal growth and the subsequent amount of freshwater that could be conserved in the underlying confined aquifer systems.

Corn and soybeans were the row crops selected for evaluation in this study as they constitute the major non-food crops grown in Maryland and can be supplemented for irrigation from WWTPs (recycled water) as the water source (MDE 2020; Shoushtarian & Negahban-Azar 2020). The two scenarios were defined as:

Scenario 1: There was no irrigation management included – the model is run for crop yield under rainfed conditions.

Scenario 2: Irrigation was included in this scenario – the model is run for crop yield using an auto-irrigation subroutine that supplements the plant water deficit from an outside source, interpreted as the nearby WWTPs. The irrigation is triggered by water stress, which is based on the plant water demand. This water stress is evaluated within the SWAT model as per the water balance equation.

The complete framework of the study is presented in Figure 2, including the required inputs for model development, model outputs that were evaluated during the simulation, and the scenarios that were evaluated.
Figure 2

The modeling framework for the study.

Figure 2

The modeling framework for the study.

Close modal
Planting, harvest, and auto-irrigation scheduling

This research integrated row crop management practices, specifically planting and harvesting, based on the heat units' approach. This approach was employed to accommodate the climatic variability affecting the onset and conclusion of the growing season throughout the 25-year simulation period.

Irrigation scheduling for the irrigation management scenario (Scenario 2) was incorporated using the auto-irrigation scheduling available in SWAT and triggered based on plant water stress. In this case, the water stress is based on plant water demand. Here, the water stress threshold is a fraction of potential plant growth. Anytime actual plant growth falls below this threshold fraction due to water stress; the model will automatically apply water. As per the availability of water in the irrigation source, the model will add water to the soil until it is at the field capacity. The plant water stress threshold to trigger irrigation was assigned a value of 0.65. As corn and soybeans can tolerate a maximum of 50% of water stress (Rhoads & Yonts 1991; Staton 2020), 65% of water stress was selected as the trigger point of irrigation. Irrigation efficiency was set as 0.6, irrigation surface runoff ratio, which is the fraction of irrigated water that is converted to runoff, was set as 0.1, and the maximum irrigation depth for corn and soybeans at each trigger was set at 25 mm (Rhoads & Yonts 1991; Staton 2020).

Evaluation of different climatic conditions

The two irrigation management scenarios were further evaluated for multiple years with varying precipitation amounts during the growing season to evaluate climate impacts (i.e., dry versus wet hydrologic conditions) on crop yield and crop water requirements. This was vital to understand the differences in crop water requirement during drought vs non-drought growing seasons and assess the capabilities of the WWTPs in the study region to provide supplemental water for irrigation during such periods. The changes in crop yield with supplemental irrigation from WWTPs were also evaluated to demonstrate the potential benefits of supplementing irrigation from WWTPs, especially in drought years.

Three representative years were selected based on the growing season precipitation amount. These included: (1) average precipitation year – in which the rainfall amount in the growing season is close to the average growing season rainfall of the whole simulation period, (2) wet year – in which the growing season rainfall is higher than the average growing season rainfall, and (3) dry years – in which the growing season rainfall is lower than the average growing season rainfall.

To select these representative years for evaluating crop yield and water requirement for different climatic conditions, monthly precipitation during the growing season for each year (May–September) was compared against average values for the whole evaluation period (1998–2020) (Figure 3). It was identified that the years 2003 (total of 887 mm for the growing season), 2007 (277 mm), and 2009 (662 mm) represented a wet year, a dry year, and an average precipitation year of precipitation, respectively, when compared to the average precipitation during the growing season for the whole simulation period (595 mm). The monthly average value of precipitation during the growing season was 120 mm for the whole simulation period, while for years 2003, 2007, and 2009, the average growing season precipitation values were 177, 130, and 55 mm, respectively. Figure 3 shows the total monthly precipitation and average monthly precipitation during the growing season for the total simulation period compared against the years selected as representative of a wet, dry, and average precipitation year.
Figure 3

Monthly precipitation during the growing season of years 1998–2022 for average precipitation (2009), dry (2007), and wet (2003) hydrologic years. Blue color shows all-year average values; orange, gray, and yellow colors show precipitation of wet, average precipitation, and dry years, respectively. Lines show the monthly average precipitation of the growing season for respective years with similar colors.

Figure 3

Monthly precipitation during the growing season of years 1998–2022 for average precipitation (2009), dry (2007), and wet (2003) hydrologic years. Blue color shows all-year average values; orange, gray, and yellow colors show precipitation of wet, average precipitation, and dry years, respectively. Lines show the monthly average precipitation of the growing season for respective years with similar colors.

Close modal

Model calibration and validation

Crop yield calibration and validation

The SWAT model parameters calibrated for corn and soybean yield along with the default and calibrated values are presented in Table 1.

Table 1

List of SWAT model's crop parameters along with their default and SWAT-CUP calibrated values for corn and soybeans yield with respect to the NASS observed yield data (rainfed)

ParametersParameter definitionCorn
Soybeans
DefaultCalibratedDefaultCalibrated
BIO_E Biomass/energy ratio 39 36 25 15 
VPDFR Vapor pressure deficit corresponding to the fraction of maximum stomatal conductance 2.94 2.2 
GSI Max stomatal conductance (in drought conditions) 0.007 0.7 0.007 0.6 
WAVP Rate of decline in radiation use efficiency per unit increase in vapor pressure deficit 7.2 
HVSTI Harvest index 0.5 0.542 0.31 0.48 
WSYF The lower limit of the harvest index 0.3 0.16 0.01 0.01 
BLAI Fraction of the plant growing season corresponding to the 1st Point on the optimal leaf area development curve 3.75 
RDMX Maximum root depth 1.7 0.7 
T_OPT (°C) The optimal temperature for plant growth 25 26 25 25 
T_BASE (°C) Base temperature for plant growth 10 
ParametersParameter definitionCorn
Soybeans
DefaultCalibratedDefaultCalibrated
BIO_E Biomass/energy ratio 39 36 25 15 
VPDFR Vapor pressure deficit corresponding to the fraction of maximum stomatal conductance 2.94 2.2 
GSI Max stomatal conductance (in drought conditions) 0.007 0.7 0.007 0.6 
WAVP Rate of decline in radiation use efficiency per unit increase in vapor pressure deficit 7.2 
HVSTI Harvest index 0.5 0.542 0.31 0.48 
WSYF The lower limit of the harvest index 0.3 0.16 0.01 0.01 
BLAI Fraction of the plant growing season corresponding to the 1st Point on the optimal leaf area development curve 3.75 
RDMX Maximum root depth 1.7 0.7 
T_OPT (°C) The optimal temperature for plant growth 25 26 25 25 
T_BASE (°C) Base temperature for plant growth 10 

Graphical comparison between simulated and observed yield for corn (Figure 4 – upper row) and soybeans (Figure 4 – lower row) shows that SWAT's simulations (green bars) mimicked the observed yield (gray bars) under rainfed conditions well for the whole simulation period. The yellow stacked area in the plot shows the total precipitation of the growing season for consecutive years. The total precipitation indicates the wetness or dryness of the growing season for respective years.
Figure 4

Simulated (green color bars) and NASS observed (gray color bars) corn (upper row) and soybeans (lower row) yield values during calibration (1998–2007) and validation (2008–2015) periods for scenario 1: rainfed scenario. The yellow color stacked area shows the total growing season precipitation of consecutive years.

Figure 4

Simulated (green color bars) and NASS observed (gray color bars) corn (upper row) and soybeans (lower row) yield values during calibration (1998–2007) and validation (2008–2015) periods for scenario 1: rainfed scenario. The yellow color stacked area shows the total growing season precipitation of consecutive years.

Close modal

The calibrated model was able to reasonably replicate the observed yield for corn in 2007, which is a dry year (3,107 kg/ha observed vs 3,847 kg/ha simulated), as well as in 2003, which is a wet year (7,841 kg/ha observed vs 8,440 kg/ha simulated), thus indicating the capability of the model to accurately replicate water stress conditions sustained by the crops and vary crop yields accordingly. Similar results were also observed for soybeans with the model capturing the temporal variability and reasonably replicating the observed yield in a dry (1,203 kg/ha observed vs 975 kg/ha simulated) as well as a wet year (2,060 kg/ha observed vs 2,356 kg/ha simulated). The considerable difference observed in corn and soybean yields for wet years when compared to dry years also demonstrates the importance and benefits of supplemental irrigation in improving crop yield, thus contributing to farmers' profitability in the region.

Evaluation of statistical measures also showed that the model performed well for the simulation of corn and soybean yield during both calibration and validation periods. The statistical parameters R2, NSE, and KGE values for simulating corn yield during the calibration period were greater than 0.65 (0.75, 0.71, 0.74, respectively) and the PBIAS value was less than 10 (7.3). The values of R2, NSE, KGE, and PBIAS of corn yield for the validation period were 0.78, 0.70, 0.67, and 7.6, respectively. Both calibration and validation indicated a good model fit for corn yield. Evaluation of soybeans simulation showed a satisfactory model performance during calibration with R2, NSE, and KGE of 0.55, 0.43, and 0.75, respectively, and PBIAS of −6.6. The model, however, had difficulty replicating the observed soybeans yield during the validation period with low R2, NSE, and KGE values (0.15. 0.11, and 0.2, respectively), although the PBIAS value of −2.2 (less than 10) indicated a good fit. It should be noted that only one variety of crop (corn and soybeans) was used for the whole calibration and validation period, while the crop varieties are always improving to generate better yields under increasing heat and water. Model simulated corn and soybean yields were comparable during the calibration period (6,701 kg/ha observed vs 6,843 kg/ha simulated for corn; 1,851 kg/ha observed vs 1,836 kg/ha simulated for soybeans). However, the model under-simulated during the validation period for both crops (7,343 kg/ha observed vs 7,146 kg/ha simulated for corn; 2,110 kg/ha observed vs 1,805 kg/ha simulated for soybeans). This likely explains the model's difficulty replicating the observed yield for both corn and soybeans during the validation period.

Streamflow calibration and validation

The SWAT model parameters calibrated for flow along with the initial range and adjusted values are presented in Table 2.

Table 2

List of SWAT model's streamflow parameters along with their initial ranges and SWAT-CUP adjusted values by flow calibration with respect to the USGS observed streamflow at Zekiah watershed

ParametersParameter definitionInitial rangeAdjusted
CN2 (r) Curve number for moisture condition II −0.1 to 0.5 0.03 
ALPHA_BF(v) Baseflow recession constant (days) 0.01 to 1.0 0.73 
GW_DELAY (v) Groundwater delay (days) 10 to 450 74.9 
GWQMN (v) Threshold depth of water in the shallow aquifer required for return flow to occur (mm) 0.01 to 1,000 457.5 
GW_REVAP (v) Groundwater ‘revap’ coefficient 0.02 to 0.5 0.28 
REVAP_MN (v) Threshold depth of water in the shallow aquifer for ‘revap’ to occur (mm) 0 to 450 149.6 
SOL_AWC (r) Available water capacity of the soil layer −0.7 to 0.7 0.24 
CH_N2 (v) Manning's ‘n’ value for the main channel −0.01 to 0.3 0.12 
SOL_K (r) Saturated hydraulic conductivity −0.7 to 0.7 0.57 
CH_K2 (v) Effective hydraulic conductivity in main channel alluvium 0 to 400 241 
EPCO (v) Plant uptake compensation factor 0.01 to 0.8 0.17 
ESCO (v) Soil evaporation compensation factor 0 to 1.0 0.38 
HRU_SLP (v) Average slope steepness 0 to 0.8 0.018 
ParametersParameter definitionInitial rangeAdjusted
CN2 (r) Curve number for moisture condition II −0.1 to 0.5 0.03 
ALPHA_BF(v) Baseflow recession constant (days) 0.01 to 1.0 0.73 
GW_DELAY (v) Groundwater delay (days) 10 to 450 74.9 
GWQMN (v) Threshold depth of water in the shallow aquifer required for return flow to occur (mm) 0.01 to 1,000 457.5 
GW_REVAP (v) Groundwater ‘revap’ coefficient 0.02 to 0.5 0.28 
REVAP_MN (v) Threshold depth of water in the shallow aquifer for ‘revap’ to occur (mm) 0 to 450 149.6 
SOL_AWC (r) Available water capacity of the soil layer −0.7 to 0.7 0.24 
CH_N2 (v) Manning's ‘n’ value for the main channel −0.01 to 0.3 0.12 
SOL_K (r) Saturated hydraulic conductivity −0.7 to 0.7 0.57 
CH_K2 (v) Effective hydraulic conductivity in main channel alluvium 0 to 400 241 
EPCO (v) Plant uptake compensation factor 0.01 to 0.8 0.17 
ESCO (v) Soil evaporation compensation factor 0 to 1.0 0.38 
HRU_SLP (v) Average slope steepness 0 to 0.8 0.018 

*r, adjusted by using a relative fraction of the default value. *v, adjusted by replacing the default value.

Assessment of streamflow simulation showed that the model was able to adequately simulate the observed streamflow with R2, NSE, and KGE values of greater than 0.67 and PBIAS less than 5 during both calibration and validation periods, values which are considered satisfactory for model simulation of flow (Chu & Shirmohammadi 2004; Abbaspour et al. 2015; Paul & Negahban-Azar 2018). The p-factor and r-factor during calibration were 0.76 and 1.61, respectively, also indicating a good model fit. Figure 5 shows the observed (orange scatter plot) and best-simulated streamflow (blue line) by SWAT, along with the 95 PPU (gray stacked area). Results depicted in Figure 5 show that the model adequately simulated the observed streamflow during the calibration (upper row) and validation (lower row) periods with most observations in the 95 PPU. It can also be observed that the model did a good job of replicating the low flows as well as most of the high flows during both calibration and validation periods. The model, however, failed to capture all the low flows (e.g., October 2010) and high flows (e.g., December 2009, March 2011, September 2011, and December 2012) during the validation period. Previous studies have suggested that SWAT can have difficulty in replicating very high flows as the model runs at a daily time step and can see improvement in model performance for capturing high flows if run at a sub-daily time step (Brighenti et al. 2019).
Figure 5

Comparison between SWAT-simulated (blue line) and USGS-observed (orange scatter plots) streamflow during calibration (years 1998–2004): upper row and validation period (years 2007–2013): lower row. The gray stacked area shows 95 PPU of simulated streamflow.

Figure 5

Comparison between SWAT-simulated (blue line) and USGS-observed (orange scatter plots) streamflow during calibration (years 1998–2004): upper row and validation period (years 2007–2013): lower row. The gray stacked area shows 95 PPU of simulated streamflow.

Close modal

Scenario evaluation (rainfed vs irrigated)

Crop yield

Figure 6 shows the watershed average corn and soybeans yield for the Zekiah Watershed under scenario 1: rainfed (green bars) and scenario 2: auto-irrigation (blue bars) management scenarios and the growing season precipitation amount with the yellow stacked area. Results showed that the average corn yield (Figure 6, upper row) increased from 7,900 to 8,500 kg/ha with the incorporation of supplemental irrigation (scenario 2). Additionally, evaluation of only drought years showed that corn yield increased by as much as 4,000 kg/ha with the incorporation of supplemental irrigation through auto-irrigation (scenario 2) when compared to rainfed conditions (scenario 1). It was interesting to note that corn yield was slightly higher in years even with high precipitation (e.g., 2005, 2006, 2008, 2009, 2018, 2020) indicating the possibility that the crop could have suffered from mild water stress during the growing season under rainfed conditions which was overcome by the application of auto-irrigation resulting in higher corn yield.
Figure 6

SWAT-simulated corn (upper row) and soybeans (lower row) yield under scenario 1: rainfed condition (green bars) and scenario 2: auto-irrigation condition (blue bars). The yellow stacked area shows the total growing season precipitation of the consecutive years.

Figure 6

SWAT-simulated corn (upper row) and soybeans (lower row) yield under scenario 1: rainfed condition (green bars) and scenario 2: auto-irrigation condition (blue bars). The yellow stacked area shows the total growing season precipitation of the consecutive years.

Close modal

Like corn yield, soybeans yield (Figure 6, lower row) for the Zekiah watershed under auto-irrigation (scenario 2) also showed improvement when compared to the rainfed scenario (scenario 1). Soybean yield, over the whole simulation period (1998–2020), increased by an average of 575 kg/ha per year from 1,800 kg/ha under rainfed (scenario 1) to 2,438 kg/ha under auto-irrigation (scenario 2). The increase in yield, however, was much greater by up to 1,400 kg/ha per year when evaluated only for the drought years. It was also observed that soybean yield under auto-irrigation (scenario 2) increased slightly even in years with high precipitation when compared to scenario 1, like what was observed for corn. This also indicates that mild water stress is experienced by soybeans under rainfed conditions even in high precipitation years, thus showing the potential benefit of irrigation even in wet years.

Crop yield and subsequent water demand between the two scenarios were further evaluated in more detail for three specific years identified as years with average precipitation (2009), wet year (2003), and dry year (2007) during the growing season (as detailed in section 2.3.3.2). This evaluation helps to evaluate the variability in crop yield spatially and temporally as here the irrigation demand for the different precipitation years was determined (i.e., normal, wet, and dry hydrologic years).

A comparison of crop yield under rainfed (scenario 1) and auto-irrigation (scenario 2) for the dry year (2007) shows a considerable increase in average yield for both corn and soybeans under scenario 2. Corn yield increases from an average of 3,850 kg/ha under rainfed conditions to an average of 8,300 kg/ha under auto-irrigation. A similar increase was also observed for soybeans with average yield increasing from about 1,000 kg/ha under scenario 1 to about 2,400 kg/ha under scenario 2. Improvement in crop yield with the implementation of auto-irrigation (scenario 2) was observed for both crops even in the sample wet year (2003) and in the sample average precipitation year (2009). Corn and soybeans yield increased from 7,880 and 2,127 kg/ha under rainfed to 8,770 and 2,500 kg/ha under auto-irrigation in 2009 (average precipitation year), respectively. This increase in yield indicates the potential benefit of supplemental water via irrigation to reduce water stress and improve crop yield.

Irrigation requirement

The annual amount of water used for irrigating each row crop HRU under the auto-irrigation scenario (scenario 2) for corn and soybeans is presented in Figure 7. Spatial evaluation of water requirement over the whole watershed showed that water requirement was highest in the eastern, western, and southern parts at the edges of the watershed due to the dominance of agricultural land use in these regions of the study watershed. Of note is that the area in the central parts of the watershed is mostly forested and wetland, while the northeastern part is mostly dominated by urban land use (Figure 1).
Figure 7

Required irrigation water (mm) under wet (2003, left column), dry (2007, middle column), and average precipitation (2007, right column) hydrologic conditions for corn (top row) and soybeans (bottom row).

Figure 7

Required irrigation water (mm) under wet (2003, left column), dry (2007, middle column), and average precipitation (2007, right column) hydrologic conditions for corn (top row) and soybeans (bottom row).

Close modal

It was also observed that the irrigation amount varied greatly from close to 100 mm to about 700 mm for both corn and soybeans (Figure 7). As expected, irrigation amounts were lowest in the wet year ranging from 100 to 300 mm at different locations (Figure 7). Irrigation water requirements varied between 200 and 500 mm in an average precipitation year (year 2009). The dry year required the highest water for irrigation ranging from 350 to 700 mm across the study watershed.

The assessment of regional WWTPs' ability to augment irrigation necessitated the estimation of annual irrigation volumes, which varied according to yearly precipitation levels. This analysis was critical in determining the sufficiency of recycled wastewater for irrigation across years with diverse climatic conditions.

The amount of water needed for irrigation to supplement precipitation for meeting the crop water demand during the growing season for each of the model-delineated subbasins and for each selected year is presented in Figure 8. The irrigation volume required was evaluated at a subbasin scale to identify the regions within the watershed that had the highest water demand for irrigation. The total irrigation volume for each subbasin was estimated by calculating the irrigation volume for each row crop HRU as a product of the HRU area (ha) and total irrigation depth (mm) for the year and summing for all row crop HRUs in each subbasin. The annual volume of water available through the WWTPs in the region was estimated using the discharge per day data available from each WWTP.
Figure 8

Annual irrigation volume in MCM for corn (top row) and soybeans (bottom row) needed to supplement rainfall for each subbasin for the years 2003 (wet year), 2007 (dry year), and 2009 (average precipitation year). Numbers in the map within the study area indicate the subbasin numbers delineated by the model.

Figure 8

Annual irrigation volume in MCM for corn (top row) and soybeans (bottom row) needed to supplement rainfall for each subbasin for the years 2003 (wet year), 2007 (dry year), and 2009 (average precipitation year). Numbers in the map within the study area indicate the subbasin numbers delineated by the model.

Close modal

Results indicated that the maximum irrigation volume, which was applied in the dry year (2007) was highest in the western and eastern part of the study watershed with values ranging from 0.35 to 0.44 MCM for corn and 0.44 to 0.57 MCM for soybeans (Figure 8). The highest irrigation volume for corn was observed in subbasin 29, located in the western part of the watershed while the highest volume for soybeans was observed in subbasin 8, located in the eastern part of the watershed. Irrigation volume, as expected, was lowest during the wet year (2003). But, even in years with above-average precipitation (2003) and average precipitation (2009), the highest irrigation volumes were still in the eastern and western parts of the watershed. Table 3 shows the total amount of irrigation water required for corn and soybeans for the years 2003, 2007, and 2009. Total water use for irrigation for corn and soybeans for the wet year (2003) was 1.90 MCM (0.65 MCM for corn and 1.25 MCM for soybeans), for the year with average precipitation (2009) was 3.46 MCM (1.28 MCM for corn and 2.10 MCM for soybeans), while the dry year (2007) required 6.18 MCM for irrigation with 2.70 MCM for corn and 3.48 MCM for soybeans.

Table 3

Water use for irrigation for corn and soybeans for the subbasins with highest row crop areas for the years 2003, 2007, and 2009

Subbasin NoArea (ha)
Required water volume (MCM)
Year 2003
Year 2007
Year 2009
CornSoybeansCornSoybeansCornSoybeansCornSoybeans
13 17 0.013 0.020 0.053 0.068 0.022 0.035 
 10  0.012  0.039  0.018 
28 42 0.026 0.077 0.114 0.178 0.055 0.114 
12 0.018 0.009 0.048 0.031 0.024 0.018 
30 132 0.088 0.198 0.295 0.55 0.141 0.326 
 26  0.042  0.11  0.062 
11 29 30 0.026 0.044 0.106 0.128 0.044 0.075 
12 63 103 0.075 0.154 0.255 0.427 0.132 0.242 
13  15  0.025  0.066  0.039 
14 20 35 0.015 0.037 0.073 0.123 0.031 0.070 
15 42 42.5 0.039 0.068 0.163 0.18 0.077 0.114 
16 14 19 0.011 0.025 0.048 0.075 0.02 0.039 
17 45 13 0.033 0.021 0.154 0.057 0.066 0.035 
22 126.5 46 0.057 0.062 0.405 0.176 0.172 0.114 
24 59 44 0.075 0.057 0.251 0.176 0.128 0.11 
25  63  0.097  0.224  0.136 
26 23 49 0.017 0.066 0.079 0.18 0.039 0.123 
27 59 56 0.062 0.088 0.216 0.224 0.132 0.145 
28 107 82 0.079 0.092 0.374 0.317 0.167 0.185 
31 19 38 0.013 0.057 0.062 0.145 0.031 0.097 
Subbasin NoArea (ha)
Required water volume (MCM)
Year 2003
Year 2007
Year 2009
CornSoybeansCornSoybeansCornSoybeansCornSoybeans
13 17 0.013 0.020 0.053 0.068 0.022 0.035 
 10  0.012  0.039  0.018 
28 42 0.026 0.077 0.114 0.178 0.055 0.114 
12 0.018 0.009 0.048 0.031 0.024 0.018 
30 132 0.088 0.198 0.295 0.55 0.141 0.326 
 26  0.042  0.11  0.062 
11 29 30 0.026 0.044 0.106 0.128 0.044 0.075 
12 63 103 0.075 0.154 0.255 0.427 0.132 0.242 
13  15  0.025  0.066  0.039 
14 20 35 0.015 0.037 0.073 0.123 0.031 0.070 
15 42 42.5 0.039 0.068 0.163 0.18 0.077 0.114 
16 14 19 0.011 0.025 0.048 0.075 0.02 0.039 
17 45 13 0.033 0.021 0.154 0.057 0.066 0.035 
22 126.5 46 0.057 0.062 0.405 0.176 0.172 0.114 
24 59 44 0.075 0.057 0.251 0.176 0.128 0.11 
25  63  0.097  0.224  0.136 
26 23 49 0.017 0.066 0.079 0.18 0.039 0.123 
27 59 56 0.062 0.088 0.216 0.224 0.132 0.145 
28 107 82 0.079 0.092 0.374 0.317 0.167 0.185 
31 19 38 0.013 0.057 0.062 0.145 0.031 0.097 

Evaluation of WWTPs for supplemental irrigation

The first consideration to choose WWTPs, those can be used for irrigating crop fields is the distance or proximity of the point of use. It is a critical factor in the decision-making process for water reuse in agriculture, particularly in regions lacking established water distribution infrastructure. This is due to the logistical and economic implications associated with the conveyance of treated water. Transporting water over greater distances can entail significant costs. Moreover, the energy requirements and environmental impacts of extended distribution systems can also be substantial. Hence, closer proximity reduces these barriers, enhancing the feasibility and sustainability of water reuse initiatives. Therefore, WWTPs that are close to agricultural fields with distances less than 10 km were only included in this study.

Spatial evaluation of WWTPs of the study watershed identified four publicly owned WWTPs (WWTPs 1, 2, 3, and 5) that were in proximity and could provide recycled water for irrigation (Figure 9). Three of the four WWTPs (WWTPs 1, 2, and 5) had an annual discharge capacity of 0.07 MCM per year or less while the fourth one (WWTP 3) had a capacity of 1.7 MCM per year providing a combined capacity of 1.77 MCM per year for supplementing irrigation in the case study Zekiah watershed (Figure 9). Comparison against the annual amount of water required for irrigating corn and soybeans for the three select years with varying precipitation showed that the four WWTPs can provide an adequate supply of water to meet the demand for irrigation in a wet year while only partially being able to fulfill the demand in dry years and years with average precipitation. The combined capacity of 1.77 MCM per year from the four WWTPs was almost equal to the requirement of 1.90 MCM in a wet year whereas the four WWTPs would be able to supplement about 53% of the total demand for an average precipitation year (1.77 out of 3.46 MCM required). In the dry years, however, the WWTPs would be able to supplement only 30% of the total water requirement of 6.18 MCM per year.
Figure 9

Location of the wastewater treatment plants (WWTPs) based on proximity, considered for supplemental irrigation in relation to the case study Zekiah watershed along with its name and annual capacity in million cubic meters.

Figure 9

Location of the wastewater treatment plants (WWTPs) based on proximity, considered for supplemental irrigation in relation to the case study Zekiah watershed along with its name and annual capacity in million cubic meters.

Close modal

A fifth WWTP (WWTP 4) located about 12 km west from the western edge of the watershed was not considered a viable source due to proximity issues (Figure 9). The WWTP, however, has a capacity of 105.4 MCM per year, and if considered for irrigation, would be easily able to supplement the required amount of water for irrigation even in extremely dry years.

The assessment of simulated corn and soybean yield within the case study watershed under rainfed conditions indicated that SWAT can adequately simulate the observed water stress for the two crops and adjust crop yield accordingly. This was particularly evident in dry years, where SWAT effectively simulated crop growth and yield to align with observed data. This finding can be critical to improving confidence in the use of the SWAT model to evaluate the impacts of different climatic conditions as well as management practices on crop yield including irrigation. Time series evaluation of crop yield showed that the model performed better in replicating the yield during the calibration period and the model slightly underestimated yield for both corn and soybeans during the validation period.

Assessment of the two irrigation scenarios (rainfed and irrigated scenario) showed that irrigation can be vital to improving crop yield in the study watershed, even though the region has traditionally been under rainfed production. This was especially evident in years with below-average precipitation during the growing season. Application of supplemental irrigation can improve corn and soybean yield by 110 and 130%, respectively, when compared to below-average growing season precipitation years. Additionally, the benefit of implementing supplemental irrigation was evident even in wet years and years with average precipitation with improvement in crop yield due to the mitigation of short-term water stresses with the use of auto-irrigation. Given the projected increase in climatic variability within Maryland and the broader Mid-Atlantic region, characterized by heightened precipitation fluctuations, more frequent short-term droughts, alterations in precipitation patterns, and rising temperatures (Paul et al. 2021), it becomes crucial to integrate irrigation into row crop cultivation. This measure is essential to improve farmer's profitability and thereby contribute to long-term agricultural sustainability in the region.

As the surficial aquifer systems in the Mid-Atlantic and Maryland can easily be saltwater intruded (Panthi et al. 2022), confined aquifer systems have been the primary source of water for irrigation. However, limited recharge due to the confined nature of the aquifer systems and increasing row crop acreages with irrigation from the confined system have led to considerable declines in groundwater levels (Foster et al. 2018; Shirmohammadi 2019) and these systems might not represent a long-term sustainable source of water for irrigation. Hence, it is imperative to find alternative sources of water for irrigation in the region.

Estimation of irrigation demand in the case study showed that irrigation requirements for corn in dry years ranged up to 2.7 MCM per year of water for a total of 690 ha of corn production area, which is equivalent to 390 mm of water on average. Similarly, evaluation for soybeans showed a supplemental water requirement of 400 mm, which is 3.5 MCM per year of water for 870 ha of soybean production. For an average precipitation year, additional water requirements for corn and soybeans were 185 and 241 mm, respectively, while for wet years, the additional water requirements for corn and soybeans were 95 and 144 mm, respectively. As WWTPs are widely prevalent in Maryland, these can serve as an important source of water for supplemental irrigation. Results from the case study watershed showed that the proximity of the watershed to WWTPs, estimated water demand for the target watershed, especially during dry years, and the capacity of the WWTPs need to be considered to estimate the volume of water that can be supplemented by these systems.

The use of recycled water from WWTPs for irrigation, however, needs to follow guidelines to make sure that the treated water meets the minimum water quality requirement for application in the desired crop types as well as to avoid health hazards, soil deterioration, and soil salinization. For Maryland, as per EPA guidelines, secondary treated and disinfected wastewater can be used for irrigating processed food crops and non-food crops (Ritter 2021). Likewise, secondary treated, disinfected, and filtered wastewater can be used for irrigating agricultural food crops (Ritter 2021). There are also additional practices that can be applied to improve the quality of wastewater used for supplemental irrigation including the use of longer retention times to reduce suspended sediments and associated microbes (Qadir et al. 2007). The use of supplemental irrigation from WWTPs will also be vital to improving the sustainability of confined groundwater resources in Maryland and the Mid-Atlantic and contribute to the conservation of blue water in the region.

As the climate is changing and is expected to change further in the future, it is vital to simulate and evaluate crop water demands under a range of climatic conditions and evaluate the feasibility of alternative water sources for supplemental irrigation to improve farmers' profitability to maintain agricultural sustainability. This will also be vital to preserve the limited freshwater resources and contribute to ecological vitality. As such, findings from this study can provide important insight into the crop water demand for the major row crops in southern Maryland and the Mid-Atlantic U.S. along with a demonstration of the method and initial assessment of the use of WWTPs as an alternate source of water for irrigation.

There are important limitations that need to be considered when utilizing the results from this study. It is important to note that only a single crop variety was considered for both corn and soybean during the whole simulation period of 19 years. But in practice, crop varieties are continuously improved to improve crop yield and make the crops more resistant to water, heat, and nutrient stresses. For instance, soybean crop types have changed rapidly over time, researchers found genetically engineered seed was planted on almost all soybean farms in the US from at least 2006 onward (Vaiknoras & Hubbs 2023). This could lead to an underestimation of crop yield as well as increased uncertainty for irrigation water use when evaluated for the different climatic scenarios. Another important limitation is the lack of an economic analysis that evaluates the cost of implementing water from WWTPs for irrigation including the conveyance infrastructure which could significantly influence the use of recycled water from WWTPs for supplemental irrigation. The study also did not consider the distance of WWTPs from each farm in estimating the total water demand that can be supplemented by WWTPs, but this can be an important consideration as it can influence the cost of irrigation as well as the availability of water to different farms based on distance. The estimation of crop water demand for irrigation under different precipitation conditions during the growing season was estimated using only one watershed as a case study. Evaluation across multiple watersheds could help better estimate crop water demand, which could be important to better estimate the irrigation requirements and the capacity of the WWTPs in providing the supplemental irrigation water. Other barriers to adoption including the need for cost-share programs to increase implementation of supplemental irrigation from WWTPs as well as the willingness of the end-users to accept crops irrigated with recycled water from WWTPs should also be studied.

This study estimated crop water demand for the major row crops in southern Maryland and the Mid-Atlantic U.S. under a range of climatic conditions and assessed the potential of using WWTPs as an alternative source of water for irrigation using the SWAT model. Assessment of corn and soybean simulation under rainfed conditions and comparison to observed values showed that SWAT can adequately simulate crop water stress, irrigation demand, and crop yield. This demonstrates the adequacy of the SWAT model as a tool for developing agricultural management strategies for irrigation management as well as crop productivity.

A comparison of crop yield between irrigated and rainfed scenarios for corn and soybean production in southern Maryland and the Mid-Atlantic showed that the region, which has traditionally been under rainfed conditions, can benefit greatly from the use of supplemental irrigation to improve agricultural productivity. The use of supplemental irrigation led not only to substantial improvement in corn and soybean yield in drought years but also improvement in years with average and above-average precipitation.

The study also helped estimate the irrigation amount required to supplement growing season precipitation in a range of climatic conditions (dry, average, wet years) to meet the crop water demand. As the confined aquifer system in the region has already seen a considerable reduction in groundwater levels, and the region is already witnessing changes in climatic conditions with more short-term droughts and increasing temperature, alternate sources of water are needed for irrigation. A comparison of crop water demand to the discharge capacity of WWTPs near the study watershed showed that the treated wastewater from WWTPs can be an important source of water for irrigation in the region.

This study showed a methodology and demonstrated SWAT as a tool to assess the alternate sources of water for irrigation and the development of BMPs. The proposed methodology can be applied to similar watersheds in the Mid-Atlantic U.S. as well as watersheds in other geographical regions with similar developing irrigation needs and available alternate water sources for irrigation.

The authors would like to thank the U.S. Department of Agriculture – National Institute of Food and Agriculture (USDA-NIFA, award number: 1027960) for funding the project.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.

Abbaspour
K. C.
,
Van Genuchten
M. T.
,
Schulin
R.
&
Schläppi
E.
1997
A sequential uncertainty domain inverse procedure for estimating subsurface flow and transport parameters
.
Water Resources Research
33
(
8
),
1879
1892
.
https://doi.org/10.1029/97WR01230
.
Abbaspour
K. C.
,
Johnson
C.
&
Van Genuchten
M. T.
2004
Estimating uncertain flow and transport parameters using a sequential uncertainty fitting procedure
.
Vadose Zone Journal
3
(
4
),
1340
1352
.
https://doi.org/10.2113/3.4.1340
.
Abbaspour
K. C.
,
Yang
J.
,
Maximov
I.
,
Siber
R.
,
Bogner
K.
,
Mieleitner
J.
,
Zobrist
J.
&
Srinivasan
R.
2007
Modelling hydrology and water quality in the pre-alpine/alpine Thur watershed using SWAT
.
Journal of Hydrology
333
(
2–4
),
413
430
.
https://doi.org/10.1016/j.jhydrol.2006.09.014
.
Abbaspour
K. C.
,
Rouholahnejad
E.
,
Vaghefi
S.
,
Srinivasan
R.
,
Yang
H.
&
Kløve
B.
2015
A continental-scale hydrology and water quality model for Europe: Calibration and uncertainty of a high-resolution large-scale SWAT model
.
Journal of Hydrology
524
,
733
752
.
https://doi.org/10.1016/j.jhydrol.2015.03.027
.
Ahmadi
L.
&
Merkley
G. P.
2017
Wastewater reuse potential for irrigated agriculture
.
Irrigation Science
35
(
4
),
275
285
.
https://doi.org/10.1007/s00271-017-0539-7
.
Akhavan
S.
,
Abedi-Koupai
J.
,
Mousavi
S.-F.
,
Afyuni
M.
,
Eslamian
S.-S.
&
Abbaspour
K. C.
2010
Application of SWAT model to investigate nitrate leaching in Hamadan–Bahar Watershed, Iran
.
Agriculture, Ecosystems & Environment
139
(
4
),
675
688
.
https://doi.org/10.1016/j.agee.2010.10.015
.
Arnold
J. G.
,
Srinivasan
R.
,
Muttiah
R. S.
&
Williams
J. R.
1998
Large area hydrologic modeling and assessment part I: Model development
.
Journal of the American Water Resources Association
34
(
1
),
73
89
.
https://doi.org/10.1111/j.1752-1688.1998.tb05961.x
.
Boryan
C.
,
Yang
Z.
,
Mueller
R.
&
Craig
M.
2011
Monitoring US agriculture: The US department of agriculture, national agricultural statistics service, cropland data layer program
.
Geocarto International
26
(
5
),
341
358
.
https://doi.org/10.1080/10106049.2011.562309
.
Boswell
V.
1926
The influence of temperature upon the growth and yield of garden peas
.
Proceedings of the American Society for Horticultural Sciences
23
,
162
168
.
Brighenti
T. M.
,
Bonumá
N. B.
,
Srinivasan
R.
&
Chaffe
P. L. B.
2019
Simulating sub-daily hydrological process with SWAT: A review
.
Hydrological Sciences Journal
64
(
12
),
1415
1423
.
https://doi.org/10.1080/02626667.2019.1642477
.
Chen
Y.
,
Marek
G. W.
,
Marek
T. H.
,
Brauer
D. K.
&
Srinivasan
R.
2018
Improving SWAT auto-irrigation functions for simulating agricultural irrigation management using long-term lysimeter field data
.
Environmental Modelling & Software
99
,
25
38
.
https://doi.org/10.1016/j.envsoft.2017.09.013
.
Chiarelli
D. D.
,
D'Odorico
P.
,
Müller
M. F.
,
Mueller
N. D.
,
Davis
K. F.
,
Dell'Angelo
J.
,
Penny
G.
&
Rulli
M. C.
2022
Competition for water induced by transnational land acquisitions for agriculture
.
Nature Communications
13
(
1
),
505
.
https://doi.org/10.1038/s41467-022-28077-2
.
Chu
T.
&
Shirmohammadi
A.
2004
Evaluation of the SWAT model's hydrology component in the piedmont physiographic region of Maryland
.
Transactions of the ASAE
47
(
4
),
1057
1073
.
doi:10.13031/2013.16579
.
Costantini
M.
,
Colin
J.
&
Decharme
B.
2023
Projected climate-driven changes of water table depth in the world's major groundwater basins
.
Earth's Future
11
(
3
),
e2022EF003068
.
https://doi.org/10.1029/2022EF003068
.
Dong
Y.
,
Jiang
C.
,
Suri
M. R.
,
Pee
D.
,
Meng
L.
&
Rosenberg Goldstein
R. E.
2019
Groundwater level changes with a focus on agricultural areas in the Mid-Atlantic region of the United States, 2002–2016
.
Environmental Research
171
,
193
203
.
https://doi.org/10.1016/j.envres.2019.01.004
.
Fortes
P.
,
Simoes
S. G.
,
Armada Brás
T.
&
Amorim
F.
2022
Competing water uses between agriculture and energy: Quantifying future climate change impacts for the Portuguese power sector
.
Journal of Cleaner Production
371
,
133629
.
https://doi.org/10.1016/j.jclepro.2022.133629
.
Foster
S.
,
Pulido-Bosch
A.
,
Vallejos
Á.
,
Molina
L.
,
Llop
A.
&
MacDonald
A. M.
2018
Impact of irrigated agriculture on groundwater-recharge salinity: A major sustainability concern in semi-arid regions
.
Hydrogeology Journal
26
(
8
),
2781
2791
.
Han
W.
,
Yang
Z.
,
Di
L.
&
Mueller
R.
2012
Cropscape: A web service-based application for exploring and disseminating US conterminous geospatial cropland data products for decision support
.
Computers and Electronics in Agriculture
84
,
111
123
.
https://doi.org/10.1016/j.compag.2012.03.005
.
Houser
M.
,
Gannon
J.
,
Producers
M. G.
,
Jacobs
A.
&
Thompson
L.
2023
How Will Maryland Commodity Agriculture Respond to Climate Change? Identifying Challenges and Opportunities From Interviews with Farmers and Farm Advisors
.
Huang
Z.
,
Liu
X.
,
Sun
S.
,
Tang
Y.
,
Yuan
X.
&
Tang
Q.
2021
Global assessment of future sectoral water scarcity under adaptive inner-basin water allocation measures
.
Science of the Total Environment
783
,
146973
.
https://doi.org/10.1016/j.scitotenv.2021.146973
.
Kama
R.
,
Song
J.
,
Liu
Y.
,
Hamani
A. K. M.
,
Zhao
S.
&
Li
Z.
2023
Water availability and status of wastewater treatment and agriculture reuse in China: A review
.
Agronomy
13
(
5
),
1187
.
https://doi.org/10.3390/agronomy13051187
.
Khalid
K.
,
Ali
M. F.
,
Rahman
N. F. A.
,
Mispan
M. R.
,
Haron
S. H.
,
Othman
Z.
&
Bachok
M. F.
2016
Sensitivity analysis in watershed model using SUFI-2 algorithm
.
Procedia Engineering
162
,
441
447
.
https://doi.org/10.1016/j.proeng.2016.11.086
.
Kummu
M.
,
Guillaume
J. H.
,
de Moel
H.
,
Eisner
S.
,
Flörke
M.
,
Porkka
M.
,
Siebert
S.
,
Veldkamp
T. I.
&
Ward
P.
2016
The world's road to water scarcity: Shortage and stress in the 20th century and pathways towards sustainability
.
Scientific Reports
6
(
1
),
38495
.
Lamichhane
M.
2022
Assessment of Future Water Availability and Irrigation Water Demand under Climate Change in the Kankai River Basin, Nepal
.
Journal of Earth Science & Climatic Change
13
(
9
).
https://doi.org/10.4172/2157-7617.1000639.
MDA
2014
Maryland Department of Agriculture Annual Report, 2014
.
Available from: https://mda.maryland.gov/Documents/14mda_ar.pdf (accessed 9 February 2024).
MDE
2020
Guidelines for the Use of Class IV Reclaimed Water, Maryland Department of the Environment
. .
MDP
2010
Maryland Department of Planning (MDP). 2010. Land Use/Land Cover for Maryland
.
Available from: http://www.mdp.state.md.us/OurWork/landuse.shtml (accessed 9 February 2024)
.
Minhas
P. S.
,
Saha
J. K.
,
Dotaniya
M. L.
,
Sarkar
A.
&
Saha
M.
2022
Wastewater irrigation in India: Current status, impacts, and response options
.
Science of the Total Environment
808
,
152001
.
https://doi.org/10.1016/j.scitotenv.2021.152001
.
Moriasi
D. N.
,
Gitau
M. W.
,
Pai
N.
&
Daggupati
P.
2015
Hydrologic and water quality models: Performance measures and evaluation criteria
.
Transactions of the ASABE
58
(
6
),
1763
1785
.
doi:10.13031/trans.58.10715
.
Narsimlu
B.
,
Gosain
A. K.
,
Chahar
B. R.
,
Singh
S. K.
&
Srivastava
P. K.
2015
SWAT model calibration and uncertainty analysis for streamflow prediction in the Kunwari River Basin, India, using sequential uncertainty fitting
.
Environmental Processes
2
(
1
),
79
95
.
https://doi.org/10.1007/s40710-015-0064-8
.
Neitsch
S. L.
,
Arnold
J. G.
,
Kiniry
J. R.
&
Williams
J. R.
2011
Soil and Water Assessment Tool Theoretical Documentation Version 2009
.
Texas Water Resources Institute
.
Available from: http://twri.tamu.edu/reports/2011/tr406.pdf (accessed 9 February 2024)
.
NRCS
1974
Natural Resources Conservation Service (NRCS)
.
Soil Survey of Charles County
,
MD
.
Olivieri
A. W.
,
Pecson
B.
,
Crook
J.
&
Hultquist
R.
2020
California water reuse – past, present and future perspectives
. In:
Advances in Chemical Pollution, Environmental Management and Protection
, Vol.
5
.
Elsevier
, pp.
65
111
.
https://doi.org/10.1016/bs.apmp.2020.07.002
.
Panthi
J.
,
Pradhanang
S. M.
,
Nolte
A.
&
Boving
T. B.
2022
Saltwater intrusion into coastal aquifers in the contiguous United States – A systematic review of investigation approaches and monitoring networks
.
Science of the Total Environment
836
,
155641
.
https://doi.org/10.1016/j.scitotenv.2022.155641
.
Paul
M.
&
Negahban-Azar
M.
2018
Sensitivity and uncertainty analysis for streamflow prediction using multiple optimization algorithms and objective functions: San Joaquin Watershed, California
.
Modeling Earth Systems and Environment
4
(
4
),
1509
1525
.
https://doi.org/10.1007/s40808-018-0483-4
.
Paul
M.
,
Dangol
S.
,
Kholodovsky
V.
,
Sapkota
A. R.
,
Negahban-Azar
M.
&
Lansing
S.
2020
Modeling the impacts of climate change on crop yield and irrigation in the Monocacy River Watershed, USA
.
Climate
8
(
12
),
139
.
https://doi.org/10.3390/cli8120139
.
Paul
M.
,
Negahban-Azar
M.
&
Shirmohammadi
A.
2021
Assessing crop water productivity under different irrigation scenarios in the Mid–Atlantic Region
.
Water
13
(
13
),
1826
.
https://doi.org/10.3390/w13131826
.
Pfister
S.
,
Bayer
P.
,
Koehler
A.
&
Hellweg
S.
2011
Projected water consumption in future global agriculture: Scenarios and related impacts
.
Science of the Total Environment
409
(
20
),
4206
4216
.
https://doi.org/10.1016/j.scitotenv.2011.07.019
.
Plumb
A. L.
2004
Enforcement and compliance history online (ECHO)
.
Reference Reviews
18
(
8
),
37
38
.
https://doi.org/10.1108/09504120410565873
.
Pyke
C. R.
&
Najjar
R.
n.d.
Climate Change and the Chesapeake Bay
.
Available from: https://www.chesapeake.org/pubs/Pubs/climchangereport.pdf (accessed 6 February 2024).
Qadir
M.
,
Wichelns
D.
,
Raschid-Sally
L.
,
Minhas
P. S.
,
Drechsel
P.
,
Bahri
A.
&
McCornick
P.
2007
Agricultural use of 11 Marginal-Quality Water-Opportunities and Challenges
.
Reilly
T. E.
,
Dennehy
K. F.
,
Alley
W. M.
&
Cunningham
W. L.
2008
Ground-water Availability in the United States
.
US Geological Survey Circular 1323
, pp.
70
.
https://doi.org/10.3133/cir1323
Reimer
A.
,
Doll
J. E.
,
Boring
T. J.
&
Zimnicki
T.
2023
Scaling up conservation agriculture: An exploration of challenges and opportunities through a stakeholder engagement process
.
Journal of Environmental Quality
52
(
3
),
465
475
.
https://doi.org/10.1002/jeq2.20317
.
Rhoads
F. M.
&
Yonts
C. D.
1991
Irrigation scheduling for corn: Why and how
.
National Corn Handbook
20
.
Ritchie
H.
&
Roser
M.
2017
Water use and stress. Our World in Data. Available from: https://ourworldindata.org/water-use-stress. (accessed 6 February 2024)
.
Ritter
W.
2021
State regulations and guidelines for wastewater reuse for irrigation in the U.S
.
Water
13
(
20
),
2818
.
https://doi.org/10.3390/w13202818
.
Sheikh
B.
,
Nelson
K. L.
,
Haddad
B.
&
Thebo
A.
2018
Grey water: Agricultural use of reclaimed water in California
.
Journal of Contemporary Water Research & Education
165
(
1
),
28
41
.
https://doi.org/10.1111/j.1936-704X.2018.03291.x
.
Shirmohammadi
A.
2019
Stressed Aquifers in the Coastal Plain of Maryland
. In
Chapman Conference on the Quest for Sustainability of Heavily Stressed Aquifers at Regional to Global Scales
.
AGU
.
Shoushtarian
F.
&
Negahban-Azar
M.
2020
Worldwide regulations and guidelines for agricultural water reuse: A critical review
.
Water
12
(
4
),
971
.
https://doi.org/10.3390/w12040971
.
Srinivasan
R.
,
Zhang
X.
&
Arnold
J.
2010
SWAT ungauged: Hydrological budget and crop yield predictions in the Upper Mississippi River Basin
.
Transactions of the ASABE
53
(
5
),
1533
1546
.
https://doi.org/10.13031/2013.34903
.
Staton
M.
2020
Moisture Stress and High Temperature Effects on Soybean Yields
.
Michigan State University Extension
.
Stermitz
R. J.
,
Nielsen
G.
&
Long
D.
1999
Testing Quality of Soil Survey Geographic (SSURGO) Database for Precision Farming
, pp.
319
326
.
Taylor
R. G.
,
Scanlon
B.
,
Döll
P.
,
Rodell
M.
,
Van Beek
R.
,
Wada
Y.
,
Longuevergne
L.
,
Leblanc
M.
,
Famiglietti
J. S.
,
Edmunds
M.
,
Konikow
L.
,
Green
T. R.
,
Chen
J.
,
Taniguchi
M.
,
Bierkens
M. F. P.
,
MacDonald
A.
,
Fan
Y.
,
Maxwell
R. M.
,
Yechieli
Y.
&
… Treidel
H.
2013
Ground water and climate change
.
Nature Climate Change
3
(
4
),
322
329
.
https://doi.org/10.1038/nclimate1744
.
Tibebe
M.
,
Melesse
A. M.
,
Zemadim
B.
,
2016
Runoff estimation and water demand analysis for Holetta River, Awash Subbasin, Ethiopia using SWAT and CropWat models
. In:
Landscape Dynamics, Soils and Hydrological Processes in Varied Climates
(
Melesse
A. M.
&
Abtew
W.
eds.).
Springer International Publishing
, pp.
113
140
.
https://doi.org/10.1007/978-3-319-18787-7_7
.
Uniyal
B.
,
Dietrich
J.
,
Vu
N. Q.
,
Jha
M. K.
&
Arumí
J. L.
2019
Simulation of regional irrigation requirement with SWAT in different agro-climatic zones driven by observed climate and two reanalysis datasets
.
Science of the Total Environment
649
,
846
865
.
https://doi.org/10.1016/j.scitotenv.2018.08.248
.
Urkiaga
A.
,
De Las Fuentes
L.
,
Bis
B.
,
Chiru
E.
,
Balasz
B.
&
Hernández
F.
2008
Development of analysis tools for social, economic and ecological effects of water reuse
.
Desalination
218
(
1–3
),
81
91
.
https://doi.org/10.1016/j.desal.2006.08.023
.
Vaiknoras
K.
&
Hubbs
T.
2023
Characteristics and Trends of U.S. Soybean Production Practices, Costs, and Returns Since 2002
. https://doi.org/10.32747/2023.8023698.ers.
USDA Economic Research Service
.
Wang
D.
,
Zhang
B.
,
Feng
H.
,
Fei
Y.
,
Wei
Z.
,
Di
F.
&
Tian
Y.
2023
Assessing the groundwater deficit for agriculture requirements under precipitation change while achieving food and water security in the North China Plain
.
Ecological Indicators
155
,
111038
.
https://doi.org/10.1016/j.ecolind.2023.111038
.
Williams
J. R.
,
Jones
C. A.
&
Dyke
P. T.
1984
The EPIC model and its application
. In
Proc. Int. Symposium on Minimum Data Sets for Agrotechnology Transfer
,
Patancheru, India
, pp.
111
121
.
Xia
Y.
,
Mitchell
K.
,
Ek
M.
,
Sheffield
J.
,
Cosgrove
B.
,
Wood
E.
,
Luo
L.
,
Alonge
C.
,
Wei
H.
,
Meng
J.
,
Livneh
B.
,
Lettenmaier
D.
,
Koren
V.
,
Duan
Q.
,
Mo
K.
,
Fan
Y.
&
Mocko
D.
2012
Continental-scale water and energy flux analysis and validation for the North American Land Data Assimilation System project phase 2 (NLDAS-2): 1. Intercomparison and application of model products: WATER AND ENERGY FLUX ANALYSIS
.
Journal of Geophysical Research: Atmospheres
117
(
D3
),
n/a
n/a
.
https://doi.org/10.1029/2011JD016048
.
Yang
Y.
,
Jin
Z.
,
Mueller
N. D.
,
Driscoll
A. W.
,
Hernandez
R. R.
,
Grodsky
S. M.
,
Sloat
L. L.
,
Chester
M. V.
,
Zhu
Y.-G.
&
Lobell
D. B.
2023
Sustainable irrigation and climate feedbacks
.
Nature Food
4
(
8
),
654
663
.
https://doi.org/10.1038/s43016-023-00821-x
.
Yildiz
V.
,
Hatipoglu
M. A.
&
Kumcu
S. Y.
2022
Climate change impacts water resources
. In:
Water and Wastewater Management: Global Problems and Measures
(Bahadir, M. & Haarstrick, A., eds.)
.
Springer International Publishing
,
Cham
, pp.
17
25
.
Zare
M.
,
Azam
S.
&
Sauchyn
D.
2023
Simulation of climate change impacts on crop yield in the Saskatchewan grain belt using an improved SWAT model
.
Agriculture
13
(
11
),
2102
.
https://doi.org/10.3390/agriculture13112102
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).